Adaptive design for periodic beacon control mechanisms in vanets / Syed Adeel Ali Shah

• Main Author: Syed Adeel , Ali Shah
• Format: Thesis
• Published: 2016
• Subjects: QA75 Electronic computers. Computer science
• Online Access: http://studentsrepo.um.edu.my/9308/1/Syed_Adeel_Ali_Shah.pdf; http://studentsrepo.um.edu.my/9308/6/adeel.pdf; http://studentsrepo.um.edu.my/9308/

In the past few years, the role of Information and Communication Technology (ICT) as a co-pilot for the drivers has shown potential in improving traffic safety and efficiency. The use of ICT enables the spontaneous wireless communication among vehicles, which is a fundamental requirement for vehicular safety systems. The efficiency of vehicular safety systems lends itself to the timely delivery of 1-hop periodic messages called beacons. The periodic beacons serve two main purposes: 1) to maintain local topology view, and 2) to inform drivers about the potential hazardous road/traffic conditions. It is worth mentioning that the design of dedicated short range communication (DSRC) standard for beaconing is motivated by the spontaneous ad hoc communication requirements under mobility. Alternatively, the DSRC standard is not fully compliant with the communication requirements of vehicular applications. That is, safety applications transmit beacons periodically and at high frequencies. Therefore, channel saturation and subsequent message collisions are inevitable under high scale ITS deployment, which cannot be addressed by the DSRC alone. Thus, beaconing under DSRC standard confines the accuracy of mutual topology awareness and the desired performance of vehicular safety applications. Clearly, the goal of this thesis is to propose adaptive designs for periodic beacon control mechanisms to improve mutual awareness and reliability of vehicular applications. Initially, this study analyses the existing adaptive beaconing approaches to identify challenges which are critical to address, such as fairness in congestion control, satisfying coverage requirement of applications, minimizing overall synchronous collisions and collisions from a specified vehicle. Subsequently, we consider the most germane parameters for designing adaptive control mechanisms, namely transmit power, contention window size and back-off selection mechanism. The first beaconing approach is based on transmit power adaptation, which provides fairness in selecting transmit power during congestion. It uses a novel cooperative game-theoretic approach to model the marginal contributions of vehicles and enable a proportional power decrease for every vehicle to minimize congestion. Another beaconing approach is proposed to control congestion by adapting differentiating transmit powers for different message types. Explicitly, the design gives a besteffort approach to maximize coverage for the event-driven messages. This is achieved by considering the application requirements and adapting the transmit power for periodic beacons with respect to the channel states. The problem of synchronous collisions is also tackled with a weighted contention window adaptation scheme. The proposed design replaces the aggressive behaviour of binary exponential back-off in the post-transmit phase of beacons and replaces it with a probabilistic selection of window size. In order to reduce collision in high density networks, the channel states are translated into meaningful weights for the appropriate contention window size selection. Apart from addressing the problem of overall synchronous collisions, another beaconing approach is proposed to minimize synchronous beacon collisions transmitted from a specified vehicle. This design works on the hypothesis that synchronous beacon collisions transmitted by a subject vehicle can be reduced if all of its neighbours predict and select different back-offs than the ones selected by the subject vehicle. The implementation of these approaches using a discrete-event simulation shows the practicality of the proposed approaches.